Despite the numerous AI platforms that provide routes to automatically generate functional music, many artists who have worked openly with AI have approached the concept through more individual means.
Take Holly Herndon, the Berlin-based composer and musicologist who recently created her own intelligent musical accomplice. Dubbed “Spawn,” this song sample generator learned from Herndon and partner Mat Dryhurst to reproduce a range of song types (including her own) through months of training its complex neural network. Spawn was able to organically add vocals to tracks presented to it.
Although, as Herndon told Art in America, the process is still making its way: “AI is not that smart, it is very low fidelity, it is not real-time, it is very slow and impractical. Spawn can take more than 24 hours to process a person’s vocal input. On the other hand, it has some unique capabilities that are quite exciting-slash-scary. The AI can extract and recreate the logic of something outside of the operator’s own logic. This is completely new for computer music.”
A virtual colleague
Herndon’s approach – to redirect the often-predictable creative choices of the human musician and carve out new musical avenues of exploration – is a commonality shared by numerous artists who have worked with AI. Alex Da Kid’s “collaboration” with IBM’s Watson was triggered by the Grammy-winning producer’s interest in whether it was really possible to create a song with a virtual colleague.
Watson uses an accumulation of data collected by a web of smart APIs. These include Watson Alchemy Language, which studies media for five years to determine current pop-cultural themes, Watson Tone Analyzer, which similarly analyzed about 26,000 song lyrics, and, crucially, Watson Beat, which determines the best chords, keys and frames to match. with a specific theme. By rail Not easy, the pair explored the heartbreak theme – producing a stunning statement, which also saw interjections from Wiz Khalifa, Elle King and the X Ambassadors. The brilliant end result was popular enough to top both the iTunes and Spotify charts.
While Alex and Holly’s involvement in AI has been driven by a desire to explore the potential of AI, stunning YouTube star Taryn Southern I am AI LP was created when the singer/songwriter found it difficult to realize the musical ideas she had in her head. Using a combination of AI music generators Amper Music and Aiva, as well as Google’s Magenta and IBM Watson, Taryn created a new musical toolkit to work with. As Southern told Digital trendsher approach was to use the fine-tuning tools of software like Amper to deviate from her original ideas, download the stems, and then rearrange the final results in her DAW.
These artists, each venturing further down the AI rabbit hole, albeit differently, are at the forefront of a new paradigm for creators. It’s almost a question of ‘when’ rather than ‘if’ AI is no longer considered an industry buzzword and becomes an everyday facet of music creation. Through their work, these artists lay the foundation for the fruitful incorporation of AI into the creative process.
Learn to fly
While these examples are just the surface of how AI has been applied by artists in different genres and contexts (see our list of ten AI-built records), AI in other areas has already started fully informing our daily lives.
It is completely immersed in how we listen to our music. The pervasive streaming platforms and music listening services are algorithms that cleverly serve similar songs to the ones we play regularly, putting together cleverly curated playlists based on the data provided by our listening habits.
Spotify’s ‘Discover Weekly’ smart playlists were designed in collaboration with French AI startup Niland. This smart neural network determines how best to populate these lists by scanning other users’ playlists of these songs and analyzing the waveforms of the songs to determine musicological similarities. “We’re working on a number of ways to improve the experience even further.”
Spotify’s research leader, Rishabh Mehrotra explained to: AI news“Reinforcement learning will be a key focus when looking for ways to optimize for a lifetime of satisfying content, rather than optimizing for the next stream. In a way, it’s not about giving users what they want now, but developing their tastes and looking at their long-term trajectories.”
Do not be afraid
We take the liberty of evading the numerous prophecies of doom predicted by those who fear AI invading their financial territory. Instead, let’s consider the alternative, that while online AI platforms can easily create the kind of utilitarian music that accompanies a simple YouTube video or podcast, for real soundtrack demands, real human composers will certainly be needed.
Feeling, heart, soul and also the ‘story’ of the soundtrack process are all essential to modern film, television and video game soundtracks. Also in production terms, there is so much that AI can bring, rather than take away, from the modern producer’s workflow. “Will AI take away all mix jobs? Absolutely not,” said Alexander Wankhammer, CMO of Sonible.
“Will it change the way the industry works in some areas? Certainly yes. I think people shouldn’t be afraid of the new possibilities AI is creating – they need to stay curious and see AI as a new, really exciting tool that (until to some extent) will determine the future of music production.”
10 recordings that use AI
1. Outside – David Bowie (1995)
While not strictly speaking “AI” in the modern sense, Bowie’s use of the Verbasizer software to concoct new lyrical ideas was a sort of digital continuation of the randomizing “dividing” technique he had relied on for previous works. Many of the songs on Bowie’s hidden gem Outside were mapped out based on these computer-suggested ideas.
2. Not Easy – Alex Da Kid (2016)
A collaboration between Dr. Dre and Eminem producer Alex Da Kid and IBM’s cultural learning and idea-building digital intelligence, Watson. Asked to explore the “heartbreak” theme, Watson gave Alex an appropriate rhythmic, lyrical and chord base. The stunning results also featured Wiz Khalifa, Elle King and X Ambassadors.
3. Return 0 – Ash Koosha (2018)
Created by Iranian composer Ash Koosha in collaboration with artificial intelligence Yona, Return 0 is a sometimes beautiful, sometimes challenging work that emphasizes its innovative conceit in its futuristic arrangements. Able to generate melodic sequences and even his own lyrics and vocals, we are very curious to see what Yona will do next.
4. Smart Beats – Toro Y Moi (2019)
American producer and songwriter Toro Y Moi, built in collaboration with commercially available AI soundscape designer Endel, has built a series of four tracks to promote Glaceau Smartwater. The smart beats EP contains the calm and emotional fruits of his – and his virtual assistant – work.
5. Holly Herndon – Proto (2019)
Using ‘Spawn’, Herndon’s self-trained AI vocal generator, prototypeThe fusion of Herndon’s signature artistic sensibilities and the twisting, line-bending Spawn results in a captivating listening experience. The glitchy, computerized textures compete against Holly’s creative personality in fascinating ways, ultimately leading to the aural equivalent of the merging of the two.
6. Shadow Planet – The Cotton Modules (2021)
Exploring the space between human composition and AI-driven arrangements, the debut by Oakland-based duo The Cotton Modules is a remarkable listen. In collaboration with machine learning decision maker Jukebox, the record swims in ambient, evolving textures and loops. While making the record, the duo began to see this different feeling as an instrument they had to learn to work with.
7. Hello World – SHADOW (2018)
Benoit Carré presented Sony CSL’s Flow Machines with ideas, which the AI then worked out. He took that and made a solid pop record. Instead of a neural network, Flow Machines used ‘Markov chains’, which are probability equation generators that require less data to work.
8. I’m AI – Taryn Southern (2018)
Perhaps the most talked-about AI-built record on the list. YouTube superstar Taryn Southern’s eight-track LP is made with music composition fine-tuner Amper. Southern’s work received a lot of press.
9. The Lost Ties – The 27 Club (2021)
A new idea to encourage musicians to talk openly about mental health. The 27 Club used AI to create new songs based on the music of those famous, trailblazing musicians who strangely left us all at the age of 27. With actual vocalists replacing the deceased singers, this is AI used for the human act of remembrance.
10. RATM Style Music Examples – AI Kittens (2021)
Like the The 27 Club project, this album was created by an AI who studied the typical structure of the rap metal quartet canon and generated short samples, which artists AI Kittens chopped into robust guitar workouts.